Several Optimization Algorithms Applied In Staff Scheduling System

Call center is import part between the enterprise and the customer. And as the new plat, the call center has significant sense for improving the effect and improving the competition power. The staff scheduling problem of call center is a kind of personnel schedule problem, which belongs to the operations research category. When the call center makes the l scheduling tables, it will measure requests from different aspects. Its purpose is to save the cost and improve the grade of service to achieve the biggest profit. With the raid development of the call center's industry scale and the refinement of between the skill and division of labor, restrictions of shift arrangement become increasingly complicated and the solving difficulty of staff scheduling problem become more serious than before. The administrator should ensure the service quality to satisfy the call request, and reduce the cost, which make the staff scheduling problem harder. The traditional operations research method largely cost the resources in solving the staff scheduling, but it is difficult for it to make an ideational shift arrangement schedule. With the emergence heuristic algorithm of artificial intelligence field, the solving system of staff scheduling which combines with artificial intelligence algorithm gets great progress in time and performance.The staff scheduling problem is a kind of timetable problem. The timetable problem is a representative question of those which are questions with multi-restraints, NP hard and combinatorial optimization. Regarding the NP hard question, it can not get the best optional solution theoretically, but as application question in reality we can adopt suboptimal solution instead of the optimal solution to satisfy the demand in reality. The staff scheduling problem is a kind of optimized question. Staff scheduling problem is the process of making timetable according to work plan, satisfaction of some restraints in certain time. It is of great importance to make a fair and reasonable shift arrangement which can mobilize the enthusiasm of workers, improve the work efficiency and bring great economic interests for enterprise. Staff scheduling system is composed of computer, information processing machine, knowledge database and components, which is the computer information processing system under the guide of related software.At abroad, Blue Pumpkin, Aspect are the leaders in the staff scheduling problem that provide the enterprise management software with extremely high efficiency algorithm. The concrete situation of domestic traffic center is different that of abroad. At present, the enterprises at home which own enterprise-grade traffic center solution are INtess, HollyV8 and so on. But their solutions are mainly focused on hardware solution of telephone distributor. Therefore, there is an urgent need for the Domestic enterprises are eager to gain the fair, reasonable staff scheduling system.In recent years, many scholars increase the research in random algorithm. One principle of the random algorithm kinds is according to Darwin natural evolution theory and the superior win and the inferior fail, such as Genetic Algorithm and Differential Evolution Algorithm and so on. Another random algorithm kind is the principle which is according to social animals who take their collective as power to hunt food, defend enemy and build nest. The intelligence reflected by the group and the algorithm formed during the processes of hunting and defending are defined as Particle Swarm Optimization Algorithm, Ant Colony Algorithm and so on. The merits of the random algorithm are simple use and robust performance. And other merit of random algorithm is that it has better solving ability when it refers to more restraints and lager scale problem. Based on the merits of random algorithm, it is feasible to use random algorithm to solve the shift arrangement problem.The paper mainly talks about the application of Genetic Algorithm, Differential Evolution Algorithm and Particle Swarm Optimization Algorithm in staff scheduling problem. First, I improve Genetic Algorithm, Differential Evolution Algorithm and Particle Swarm Optimization Algorithm according to the research result of random algorithms and the character of the staff scheduling problem. The problems solved by traditional Differential Evolution Algorithm and Particle Swarm Optimization Algorithm are the constrained problems in continuous domain. But staff scheduling problems is constrained problems which are in discrete domain. So the promise of using random algorithm to solve shift arrangement problems is to improve Differential Evolution Algorithm and Particle Swarm Optimization Algorithm and make them can be used to solve the constrained problems which are in discrete filter. The principle of improving the two algorithms during their processes is random. According to this principle, the operation with continuous characteristics in Differential Evolution Algorithm and Particle Swarm Optimization Algorithm will be improved.Then put the improved random algorithm into the mathematical model of shift arrangement. The effect using the improved random algorithms to solve the staff scheduling problem is better, and the scheduling result can be accept by the enterprise according to the rule data of the enterprise.After using random algorithm to solve shift arrangement, take ILOG Solver's solving effects as referential standard compare the effects of three random algorithms of solving shift arrangement. Through the effect comparison, we get the conclusion that Differential Evolution Algorithm has the highest efficiency in solving staff scheduling problem and Genetic Algorithm is inferior to it and Particle Swarm Optimization Algorithm has the lowest efficiency in shift arrangement problem. So using the DEA to solve the staff scheduling system is feasible.This paper also introduces staff scheduling system which is taken as project system with other modules: such as staff information management module, pretreatment of staff scheduling module, forecast module and results showing module. The staff information is the precondition for the staff scheduling system running. Meanwhile, pretreatment of staff scheduling module can be divided into shift class setting, shift order setting, shift group setting and shift turn rules setting. Pretreatment of staff scheduling module is the fundamental for the shift module's normal operation. In the staff scheduling system, take forecasted people needed in different period of everyday as fundamental and the results of the forecast directly influence the effects of the staff scheduling system.